TITLE: Irreducible Infeasible Subsystem (IIS) Decomposition for Probabilistically Constrained Stochastic Programming

ABSTRACT:

Probabilistically constrained stochastic programs (PC-SPs) have many applications in science and engineering but are very challenging to solve. Furthermore, linear programming (LP) provides very weak bounds on the optimal value. In this talk, we introduce a new decomposition approach using irreducible infeasible subsystem (IIS) inequalities to strengthen the LP-relaxation of PC-SPs. We first establish the theoretical results for determining IIS inequalities for the continuous case, and then extend the results to the binary case and give example illustrations. Next, we present an IIS branch-and-cut algorithm for PC-SP and report on preliminary computational results. 

Bio

Lewis Ntaimo received his Ph.D. degree in systems and industrial engineering in 2004, his M.S. degree in mining and geological engineering in 2000, and B.S. degree in mining engineering, all from the University of Arizona. He has been with Texas A&M University since 2004. Dr. Ntaimo research interests are in algorithms for large-scale stochastic optimization, systems modeling, and discrete event simulation. Recent applications include wildfire response planning, energy reduction in data centers, wind farm operations and maintenance, and patient and resource management in healthcare. His research has been funded by the National Science Foundation, Department of Homeland Security, and industry. Dr. Ntaimo is a member of INFORMS and IISE. He served as Vice Chair for the INFORMS Optimization Society from 2008-2010 and is now Vice President for the INFORMS Minority Issues Forum. He is on the Editorial Board of the Journal of Global Optimization is a member of the technical committee for the Society of Computer Simulation DEVS symposium.